Port rhetorics visualization to real RDFa

[clange]

In the first implementation of the rhetorics visualization we embedded the annotations required on the client side in a crude mixture of RDFa and microformat style because of lazyness. Now that RDFa is a W3C recommendation and the KWARC group is considering to adopt RDFa for OMDoc and related projects, we should also adopt this as a requirement for JOBAD: whenever embedded annotations/metadata are required by the JavaScript code, they should be embedded into the HTML as proper RDFa. Once we have a general implementation (#266), the JavaScript code for the rhetorics visualization needs to be adapted accordingly.

Once this is done, delete xsl/varia/rhetorical-visualization, the old XSLT, which adds interactive controls to an OMDoc document with rhetorical structures (cf. #574)

[clange]

Note: If we have subsections of sections, we need to make them proper resources in the RDFa sense, e.g. using @about.

The specification of RDFa for metadata in OMDoc is basically done; it will also be used for the rhetorical markup, the old <omtext type="..."> being a pragmatic syntactic sugar for it. This ticket is then about preserving as much as JOBAD needs of the "RDFa in OMDoc" (in the content markup on the server) as "RDFa in XHTML" (in the active document on the client).

[clange]

This is not yet on the agenda for implementation, but it relates to the guided research via the question "what knowledge representation in the document sent to the client is needed in order to enable such a service"

[clange]

After some rearrangements, the files of the old demo temporarily ended up as follows:

• omdoc/trunk/examples/testcases: the OMDoc sources
• omdoc/trunk/examples/varia: the HTML demo, and all JavaScript and CSS that it needs
• jomdoc/src/prj/style/trunk/xsl/varia: the post-processing XSLT

From there, everything should somehow be moved into JOBAD.

[clange]

@Catalin: This is really future work, but may eventually be within your responsibility.